Former NEC Director Gary Cohn: Rising oil prices will have an impact on U.S. consumption power

Gary Cohn discusses AI capex spending, the Fed's interest rate outlook, and the potential impact of the Iran war on oil prices. The article also highlights White House opposition to Anthropic's plan to expand access to its Mythos model, adding a regulatory angle to the AI debate. Overall, the piece is mostly commentary and policy-focused rather than a direct market-moving event.

Analysis

The more important market signal here is not the AI commentary itself, but the tightening link between regulatory posture and capex pacing. If policymakers start constraining access to frontier models or deployment channels, hyperscalers and enterprise software buyers will become more selective on near-term rollout, which can flatten the second derivative of AI spend even if the long-run thesis remains intact. That would pressure the “all AI spending is good spending” trade and favor infrastructure owners with existing contracted demand over pure-play model aspirants.

Energy is the cleaner second-order hedge. Any sustained oil premium from geopolitical escalation tends to hit longer-duration growth multiples first through higher discount rates and margin compression in energy-intensive software and hardware supply chains. The market often prices these shocks as transitory for 1-2 weeks, but if crude stays bid for a month, investors typically rotate toward cash-generative megacap tech with pricing power and away from names where AI capex is still pre-revenue.

On rates, the important risk is not whether cuts happen, but whether the market has become too dependent on an easing path to justify current multiple expansion. If inflation re-accelerates via energy, the front end can reprice higher even without a full policy reversal, and that disproportionately hurts the most crowded duration-sensitive AI winners. The contrarian setup is that any policy or geopolitical headline that slows AI commercialization may be bearish for speculative momentum, but mildly bullish for incumbents like IBM that monetize governance, integration, and regulated enterprise adoption rather than frontier model expansion.

Key Decisions for Investors

• Fade the most duration-sensitive AI basket on strength over the next 2-4 weeks: short a basket of high-multiple AI/software names against long IBM as a relative-value hedge if regulatory scrutiny on model access intensifies.
• Use oil-volatility as a macro hedge: buy 1-2 month calls on XLE or XOP if crude begins to reprice higher, with a stop if headlines fade and oil retraces below the prior breakout level.
• If front-end yields pop on renewed inflation/geopolitical pressure, add a tactical short in unprofitable AI software names via put spreads; target a 10-15% drawdown over 30-45 days with defined premium risk.
• Maintain a long IBM / short frontier-AI-beta pair for 1-3 months: the thesis is that enterprise AI spend shifts toward compliance-ready integration layers before it broadens to riskier model expansion.
• Avoid chasing the AI capex theme until there is confirmation that enterprise deployment budgets, not just vendor announcements, are accelerating; otherwise the risk/reward is poor after a crowded move.